Sure Start closures methodology

This page describes how FamilyPlans assembles its Sure Start closures dataset, the verification levels we use, the sources we draw on, and the process for updates and corrections. We publish this methodology before the public map and we do not publish individual map markers without an evidenced source for that location.

Verification levels

Each centre in our dataset has one of these verification statuses. The map publishes individual markers for verified_closed, transitioned, and replaced_by_family_hub only. Counts for the others are shown aggregate-only, with the methodology disclosed.

• verified_closed — we have a council press release, FOI response, or DfE record that names this specific location and its closure date.
• likely_closed — multiple consistent sources but no single primary source we'd quote on its own.
• transitioned — re-badged or repurposed (often into a Family Hub) rather than fully closed.
• replaced_by_family_hub — explicit successor record links to a current Best Start Family Hub.
• aggregate_only — we hold council-level totals (e.g. from DfE statistical releases) but no individual location records yet.
• uncertain — the centre exists in our index but we don't hold sufficient evidence to assign a status.
• disputed — a council, charity, or member of the public has challenged our classification; we re-investigate before changing.

Current dataset state

Current counts (no record updates yet):

• 0 centres classified verified_closed.
• 0 centres classified likely_closed.
• 0 centres classified transitioned.
• 0 centres classified replaced_by_family_hub.
• 0 aggregate-only records (counts at council level without individual evidence).
• 0 centres marked uncertain pending evidence.
• 0 closure / opening / transition events recorded against individual centres.

We do not publish one fixed “X centres closed since 2010” headline. Counts here update from verified records and are shown with their verification level.

Sources we draw on

• DfE published Sure Start centre data (2010-2017) and Children's Centre statistical releases.
• DfE Best Start Family Hubs published list (current).
• IFS “Sure Start: A Tale of Decline and Recovery” (2021) for council-level analysis.
• Council press releases, council web archives, and FOI responses for individual closures.
• Charity sector research (Action for Children, NCB, Save the Children) for context.
• Parliamentary research briefings from the House of Commons Library.

The full source register, with licence and review status for each entry, is at /methodology/sources.

How we get from a source to a record

1. Each source is registered, robots-checked, terms-reviewed, and assigned a refresh cadence before any scraper runs (see /methodology/sources).
2. Scrapers fetch source pages, store immutable snapshots in private object storage, and extract claims into our claims ledger with a content hash, fetched timestamp, and parser version.
3. Each centre's verification_status is set based on the strongest available evidence — we never use language stronger than the source supports.
4. Disputed records are flagged and re-investigated; corrections appear at /research/corrections.

How readers can challenge a record

If you have evidence that contradicts our classification of a specific centre, please report it with a source link. We aim to acknowledge within 2 business days and resolve within 5. Particularly welcome:

• Council communications about specific closures or transitions.
• Local press coverage we may have missed.
• Family Hub successor information for centres we've marked as closed.
• FOI responses you've received and would be willing to share.

What we do not do

• We do not publish fixed national-level closure counts without showing the underlying verification level.
• We do not classify a centre as verified_closed without a primary source for that specific centre.
• We do not show individual map markers for aggregate_only or uncertain records.
• We do not republish source PDFs in full; we link out and quote in context.
• We do not use AI extraction for closure status determinations in production.

The Sure Start closures map publishes verified records only. See /research/sure-start-closures for the interactive view.